Understanding genetic changes in cancer will lead to a better diagnosis, prognosis and therapy. In synergy with large scale cancer genomics projects, such as The Cancer Genome Altas (TCGA), a computational biology research team at Memorial Sloan Kettering Cancer Center plans to elucidate the role of thousands of protein mutations observed in many human tumor samples. The Online Mutation Assessor (OMA), a web-accessible computational resource, will combine detailed knowledge about the evolution and structure of proteins and their interactions in biomolecular pathways with advanced algorithms and software technology. The goal is to provide researchers with an increasingly accurate and informative estimate of the functional impact of mutations observed in humantumor samples. In particular, we aim to develop a next generation of algorithms and software for the purpose of assessing the functional impact of non-synonymous cancer- associated mutations. We will apply these to all known mutations and embed the resulting predictions in the web of high-quality biological information systems such as genome databases. In close collaboration with clinical researchers, we will then analyze protein mutations observed in thousands of tumor samples in tens of cancer subtypes. The resulting detailed predictions will allow ranking of mutations in terms of relative importance and greatly increase the efficiency of targeted experiments. Knowing the likely functional impact of mutations will point to the better use of existing therapies and lead to the discovery of new drug targets for pharmaceutical development. When translated to clinical practice, detailed knowledge about an individual's mutation profile will enable a personalized approach to therapy.